178 Transactions of the Royal Microscopical Society. 
what, in the light of the facts I submit, appear mistaken grounds, 
that I am the rather inclined to see how very many of them might 
be shown to glide insensibly into each other, having no rigid line 
between. But this of course would not be enough, unless the 
developmental history in all the said cases were coincident or nearly 
approximate—a matter on which I cannot speak. But when, on 
good authority, three Navicule, called separately by specific names, 
are affirmed to be capable of being considered one species, so far as 
their life-history is concerned; and when it is further shown that 
the “striation”—a permanent feature—is the same in each; in 
consideration of the fact that both the “shape” of the outline of 
the frustule and the form of the “nodules” are constantly subject 
to minute variations, it appears to be a case, fairly made out, of 
specific identity. This, however, is, comparatively speaking, a side 
issue ; the great point is, whether it is possible to make what are 
shown to be the variable features of a diatom—taken in entire 
independence of its life-history—the elements for determining its 
species ? 
There are three things in most Navicule that vary: (1) the 
general outline of the frustule, (2) the midrib as a whole (although, 
so far as my examination goes, it is much less variable than the 
outline of the frustule), and (3) the tenwity of the beading or 
striation. There is one thing (broadly speaking) which is in- 
variable—it is the character or arrangement of the beading.* 
* T write thus guardedly because I am convinced, as I have already hinted, 
that with a complete knowledge of the life-histories of all the forms of the Navi- 
cules there might be an immense reduction of what are now known as specific 
forms. ‘Take, for example, such a prominent form as NV. granulata and compare 
with it WV. latissima, N. humerosa, and N, brevis on one side, and NV, marina, N. pusilla, 
and J, carassivs on the other side, and I suspect that complete search would find 
all the intermediate forms. In like manner there can be but little question that 
in either old or living forms the intermediate steps between JV. rhomboides and 
N. cuspidata might, although perhaps with difficulty, be found. 
The transition forms between Pleurosigma fasciola and P, macrum and P. pro- 
longatum are almost at hand without searching; and I imagine that many of us 
who have examined a large number of slides or “ material” of the forms P. angu- 
latum, P. quadratum, and P. elongatum have often wondered at what point the one 
ceased and the other began. Of course if their life-histories are essentially different 
from each other, then their remarkable similarity and blending into each other 
must be regarded merely as coincidence, or in some sense “mimicry.” But if 
their developmental histories only differ enough to account for their differences 
of form, then I think such forms extremely instructive and well worthy the care- 
ful study of those who seek for direct and living proof of developmental changes 
in existing organisms: a matter much more readily discovered in minute forms, 
seen together in enormous numbers, and subject to critical investigation in their 
entirety, than in the larger and more complex organisms, which, when taken in 
the greatest numbers in which they can be found, for critical examination, are 
but, relatively, few indeed. 
In fact, therefore, in saying that even the arrangement of the beading in the 
diatoms, chiefly discussed in the above “ Note,” is permanent, I merely mean 
that it is so relatively to the other parts; but in all probability this permanence 
is only, in reality, an indefinite persistence, a feature, that is to say, immensely 
jess liable to change or vary than others. 
